Background and Overview

Before moving to expand federal protections to subjects of currently unregulated research, we should examine the present system, which has grown incrementally over a period of years. The first federal policies covering research funded by the Department of Health, Education and Welfare (now DHHS) were issued in 1966. The first congressionally mandated commission, the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (National Commission), started its work in 1974.¹⁰It produced ten reports over four years that provide the ethical foundation for the system of protections in place today. Even so, it took until 1991 for a subset of federal agencies to agree upon the Federal Policy for the Protection of Human Subjects¹¹as the core regulation governing research conducted by or under the auspices of the government.

This policy is often referred to as “the Common Rule.” The Common Rule is not followed by all federal agencies, and it is unevenly enforced by those that do.

In 1994, Dr. Robyn Y. Nishimi of the Office of Technology Assessment (OTA), testified before Congress that:

No statute… governs the general oversight of research involving Americans. Moreover, the current system, while changing incrementally, has fallen short of implementing, or did not implement at all, recommendations made between 1973 and 1982 by an ad hoc committee of DHEW, a congressional report and two congressionally mandated commissions.¹²

Research involving human subjects may be regulated by the federal government through three separate mechanisms: a) because it is sponsored by a federal office or agency subscribing to the Common Rule;

b) because an institution conducting research not sponsored by the federal government has voluntarily granted jurisdiction over the research to OPRR through a negotiated assurance; or c) because the research involves reg-ulated drugs or medical devices over which the FDA has jurisdiction. An unknown quantity of research is not regulated either because the sponsoring/conducting agency does not subscribe to the Common Rule or has not negotiated an assurance extending federal jurisdiction or because the research is privately sponsored/conducted and not subject to FDA approval.

1. The Common Rule

Summary of Common Rule Provisions.The approach of the Common Rule to regulation of human subject research is decentralized, involving negotiation of assurances by the institutions where research is conducted with federal agencies certifying that certain procedural and substantive protections will be provided. While these assurances are received and overseen by the various federal agencies, review of specific proposed experi-mental protocols and informed consent forms occurs at the local level through IRBs. Federal requirements govern the composition and activities of IRBs, but as we shall see, true oversight and accountability for the rigor and consistency of IRBs has not been attained.

Six categories of research are exempt from full IRB review under the Common Rule.¹³These review proce-dures permit research meeting specific, narrow criteria to proceed without any formal review. The six exemp-tion categories, developed with public comment and through negotiaexemp-tion and policy formulaexemp-tion involving an interagency committee over a period of ten years, offer important insight into one mechanism that might be employed to address the practical problems that could arise from broadening the scope of federal regulation.

(See below.) There are additional categories of research for which IRBs may use expedited review procedures,

on the theory that the types of research involved, like voice recordings or collection of fingernail clippings, are less intrusive and pose a low level of risk to the subject.¹⁴

Application of the Common Rule.Seventeen federal agencies that fund or conduct research subscribe to the Common Rule and thus use an approach similar to that of DHHS, the lead federal agency in this area, with the important exception that most do not have an active program for assuring compliance with applicable regulations. While there is no definitive assessment of how many federal agencies conduct or fund research on human subjects, the 1981 report of the President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research, Protecting Human Subjects,documented that 23 federal entities funded research involving human subjects.¹⁵

Dr. Nishimi of OTA testified to Congress in 1994 that:

…a definitive picture of current federal implementation and oversight of existing regulations to protect human research subjects is not available… .Currently, information from all agencies on the total number of all research grants or contracts, total funding for research and grants involving human subjects, and number of full time equivalent personnel devoted to assurance and compliance has not been collected in a coordinated or centralized fashion.…

For some agencies, information even limited to the number of, funding levels for, and types of research involved for currentgrants or contracts using human subjects could not be reported as recently as March 1994, although the common rule has been effective since June 1991.

Without such information, ensuring that proper institutional assurances are in place and then overseeing compliance would appear to be problematic.¹⁶

Within DHHS, OPRR assumes oversight responsibility for both human and animal subjects of research.

The FDA also has responsibility for protecting the rights and welfare of human subjects of research, in the context of its required approvals for drugs and medical devices. While both OPRR and FDA have mechanisms for reviewing cases of alleged noncompliance with federal regulations and responding to them,¹⁷most other agencies do not. As Dr. Nishimi noted in 1994:

…agencies will not be aware of violations of existing regulations unless a rigorous system is in place to monitor compliance. Put another way, those Departments and agencies that are not looking for problems will not find any problems.¹⁸

The ACHRE inventoried federal experiments on human subjects and found that:

In most federal agencies, current mechanisms of oversight of research involving human sub-jects are limited to audits for cause and a review of paperwork requirements. These strategies do not provide a sufficient basis for assuring that the present system is working properly.¹⁹ 2. OPRR Oversight System

OPRR relies heavily upon the assurances it negotiates with institutions conducting research. These assurances contain the institutions’ provisions for protecting the welfare of human subjects and generally follow common patterns. In addition to the promises institutions provide in their negotiated assurances, OPRR provides educa-tional support and information to IRBs and queries institutions about reports of noncompliance. OPRR conducts a number of record reviews through paper correspondence and a much smaller number of on-site, for-cause reviews of IRB effectiveness. Both the GAO and the DHHS OIG compliment the effectiveness of OPRR’s compliance reviews, but both also comment upon the extraordinarily limited extent of on-site visits, due to staffing and budgetary constraints.

Currently, OPRR negotiates an assurance with each institution that receives research support from DHHS.²⁰ Each assurance requires significant amounts of time and review by OPRR. According to the GAO, in 1996 OPRR had about 14 full-time equivalent staff devoted to human subject protection, with a budget for those activities of under $1 million. OPRR augments its professional staff with three physician volunteers.²¹

Most major institutions accepting federal research funding negotiate Multiple Project Assurances (MPAs) with OPRR through which they agree to provide the same protections to all subjects of research conducted at the institution that they do for research funded by DHHS. There are almost 450 MPAs covering more than 750 entities operating around 700 IRBs; they are virtually all in the United States.²²Two to four times as many institutions negotiate only Single Project Assurances (SPAs) for each individual project funded in whole or in part by DHHS (covering around 3,000 IRBs), or Cooperative Project Assurances (for multisite clinical trials), with another 1,250 associated IRBs.

There are about 3,000 active SPAs— locations where we know that some DHHS-regulated research is con-ducted but no MPA is in place to cover other research that may be performed at that institution.²³At these institutions, other research involving human subjects may occur without any governmentally provided protec-tions for the subjects of that research. This does not necessarily mean that the research is not reviewed by an IRB, as institutions may choose voluntarily to extend those protections to all subjects of research— or they may not. It does mean that there is no federal jurisdiction to investigate if a subject of research files a complaint.

OPRR reports that negotiation of assurances for SPAs requires more time than other negotiations, because they usually involve OPRR scrutiny of protocols and informed consent documents from institutions with little or no history of review of research involving human subjects. Because DHHS funds research in 80 countries around the world, institutions negotiating SPAs are not all in the United States.

These and other recent reviews of the IRB system emphasize the changes that time and resource constraints have brought to their oversight by OPRR. While all contribute to a conclusion that OPRR does a good job of protecting human subjects of research, they also illustrate that its resources are inadequate for its present responsibilities and indicate areas where changes could strengthen its performance.²⁴

The assurance negotiation process, for example, has by most accounts become routinized.²⁵The NBAC-commissioned paper by Dr. McCarthy provides background on the educational nature of the assurance negotiation process in its early phases: He implies that these negotiations were usually conducted on-site at institutions and describes how mutually beneficial these exchanges were, both for institutions with little background in these issues and for OPRR officials in gaining insight into the institution’s culture. By now, the negotiation process has lost much of this educational flavor; perhaps its time has just passed.

The McCarthy paper also describes an OPRR that was able to sustain a much larger educational program than is now the case. As an ongoing constant educational program is essential if consistency is to be achieved in a decentralized system, this is a serious matter. It is not an overstatement to suggest that, in a large distrib-uted oversight system, high-quality educational programs are the cornerstone of true accountability. The report of the ACHRE went so far as to recommend that:

…efforts be undertaken on a national scale to ensure the centrality of ethics in the conduct of scientists whose research involves human subjects.… The necessary changes are unlikely to occur solely through the strengthening of federal rules and regulations or the development of harsher penalties.… The federal government must work in concert with the biomedical research community to exert leadership that alters the way in which research with human subjects is conceived and conducted so that no one in the scientific community should be able to say “I didn’t know” or “nobody told me” about the substance or importance of research ethics.²⁶

Much of OPRR’s ability to conduct such programs has since been curtailed by budgetary reductions and limitations, although an ongoing set of programs is offered annually through co-sponsorship arrangements.

Dr. McCarthy raises cautions against the conflicts of interest that can arise when regulated institutions are assuming responsibility for part of the cost of educational programs in this way.

OPRR’s reliance upon a paper-based and time-intensive assurance negotiation system is no longer desirable.

OPRR agrees with the calls from external observers that it is time to make changes in the negotiation of assur-ances.²⁷Replacing the assurance system with a streamlined registration system seems a sound alternative. If change of this nature were adopted expeditiously, it would free some resources for activities more conducive to true accountability. OPRR should be able to make this change without regulatory modification, but should be encouraged to do so by NBAC.

Other recommendations of the OIG— several of which mirror changes OPRR staff have indicated they would like to adopt— will require more resources than are presently available to OPRR. This is a central issue with which NBAC must grapple as it formulates it recommendations.

3. FDA Oversight System

The FDA is responsible for the safety and effectiveness of medicines and medical devices. As part of its regula-tory responsibilities, FDA requires that studies involving investigational new drugs, devices, and biologics receive review and approval by an approved IRB and that researchers submit statements that they will uphold ethical standards. FDA has “concurred” with the Common Rule, but has not adopted it in its entirety; while its regulations are largely congruent with those that OPRR enforces, there are differences in its IRB and informed consent regulations.

A major difference is that FDA does not require or negotiate assurances with institutions. It oversees IRBs through an inspection program, in which it routinely performs on-site procedural reviews of IRBs to determine whether they are in compliance with their own procedures and with applicable FDA regulations. The GAO reported that FDA employed about 13 full-time equivalent staff members devoted to IRB inspections in fiscal year 1995.²⁸FDA also has monitoring activities for individual drug studies and for clinical trials. Each involves reviewing compliance with consent requirements and other human subject protection protocols.

The GAO reviews concluded that while the FDA program is rigorous and that it detects (and corrects) problems in human subject research, “FDA’s inspection program is geared more toward protecting the eventual consumer of the drug than the subjects on whom the drug was tested.”²⁹If NBAC wishes to assure protection for human research subjects, this observation should trigger serious examination and consideration.

4. Nonsubscribing Federal Agencies

Subjects of research conducted or funded by federal agencies that do not subscribe to the Common Rule do not receive its core protections. There are indications that research is funded or conducted by the Nuclear Regulatory Commission, the National Endowment for the Humanities, and the Department of Labor.³⁰

In 1995, the ACHRE found that the magnitude of research conducted by federal agencies not in compliance with the Common Rule is a significant concern and recommended that there be an assessment of the level of that research. It further recommended action to “ensure that all subjects are afforded the protections it offers.”³¹ Anticipating the ACHRE findings, President Clinton issued an Executive Memorandum in 1994 intended to address gaps in government coverage; specifically, he ordered that all federal agencies and departments should come into compliance with the Common Rule and to suspend noncompliant experiments immediately.³²There is no evidence that any department or agency suspended a single activity following the President’s instruction.

The staff of NBAC is researching the issue of federal agency compliance with the Common Rule and this Executive Memorandum.

5. A Caveat— Not All Unregulated Research Goes Without Review

It is important to note that federal regulation is neither the only mechanism through which research is inde-pendently reviewed nor is it the only way participants in research are offered the protection of informed con-sent. It may not be appropriate to assume that expanding the scope of federal regulation is the only way to achieve the twin goals of assuring informed consent by subjects and objective review of protocols. Many uni-versities extend to nonfederally funded research the same oversight required by federal regulation, mandating that all research conducted at the institution is subject to review by an IRB. Of course, in virtually all cases this voluntary extension lacks independent compliance oversight, so NBAC must confront the degree to which it considers compliance oversight to be essential to a federal protection system.